Deep variational autoencoders for breast cancer tissue modeling and synthesis in SFDI

Abstract: Extracting pathology information embedded within surface optical properties in Spatial Frequency Domain Imaging (SFDI) datasets is still a rather cumbersome nonlinear translation problem, mainly constrained by intrasample and interpatient variability, as well as dataset size. The B-variational autoencoder (B-VAE) is a rather novel dimensionality reduction technique where a tractable set of latent low-dimensional embeddings can be obtained from a given dataset. These embeddings can then be sampled to synthesize new data, providing further insight into pathology variability as well as differentiability in terms of optical properties. Its applications for data classification and breast margin delineation are also discussed.

Otras comunicaciones del congreso o articulos relacionados con autores/as de la Universidad de Cantabria

 Autoría: Pardo A., López-Higuera J., Pogue B., Conde O.,

 Congreso: European Conference on Biomedical Optics, ECBO (2019 : Múnich, Alemania)

Editorial: The Optical Society (OSA) - SPIE Society of Photo-Optical Instrumentation Engineers

 Fecha de publicación: 11/07/2019

Nº de páginas: 3

Tipo de publicación: Comunicación a Congreso

DOI: 10.1117/12.2527142

ISSN: 0277-786X,1996-756X

Proyecto español: TEC2016-76021-C2-2-R

Url de la publicación: https://doi.org/10.1117/12.2527142